[v4,00/10] Add CPU-type to topology

Message

Pawan Gupta Sept. 30, 2024, 2:47 p.m. UTC

v4:
- Series doesn't apply to upstream cleanly anymore, rebased to v6.12-rc1,
  resolved the merge conflict in files:
	arch/x86/include/asm/cpu_device_id.h
	arch/x86/kernel/cpu/common.c
- Remove redundant "hw_" in intel_hw_native_model_id().

v3: https://lore.kernel.org/r/[e-mail]
- Add a patch to prepend "0x" to hex values in cpu_debug_show() (Borislav).
- Add support to to also get Intel Native Model ID (Dapeng).
- Keep similar models together in the affected processor list (Josh).
- Add a comparison of .init.data in commit message for cpu_vuln_blacklist
  before and after decluttering patch for affected processor list (Josh).
- Drop the debugfs file reference in the commit message (Borislav).
- s/cpu_type/hw_cpu_type/ (Borislav).
- Add a union for hw_cpu_type for easy retrieval of intel_core_type and
  intel_native_model_id.
- Updated commit messages, collected tags.
- Rebased to v6.11-rc3.

Note, I will be off work till Tuesday, will reply to any comments then.

v2: https://lore.kernel.org/r/[e-mail]
- Move CPU-type to the end of the CPU topology structure (Andrew).
- Use c->cpuid_level instead of cpuid_eax(0) (Andrew).
- Move CPU-type enum out of ifdef CONFIG_NUMA (kernel test robot).
- Rename cpu_type to hw_cpu_type (Borislav).
- Explain replacing get_this_hybrid_cpu_type() with topology_hw_cpu_type()
  in the commit message (Dave).
- Fix the alignment in cpu_vuln_whitelist (Andrew).
- Add the obj compare note in the commit message (Dave/Tony).
- s/X86_CPU_TYPE_INTEL_ATOM/ATOM/ in cpu_vuln_whitelist (Dave).

v1: https://lore.kernel.org/r/[e-mail]

Hi,

This series adds support for CPU-type (CPUID.1A.EAX[31-24] on Intel) to
differentiate between hybrid variants P+E, P-only, E-only that share the
same Family/Model/Stepping. One of the use case for CPU-type is the
affected CPU table for CPU vulnerabilities, which can now use the CPU-type
to filter the unaffected variants.

* Patch 1 fixes a minor formatting issue in cpu_debug_show().

* Patch 2 adds hardware cpu-type to CPU topology structure and introduces
  topology_hw_cpu_type().

* Patch 3-5 replaces usages of get_this_hybrid_cpu_type() with
  topology_hw_cpu_type().

* Patch 6-8 Updates CPU-matching infrastructure to use CPU-type.

* Patch 9 cleans up the affected CPU list.

* Patch 10 uses the CPU-type to exclude P-only parts from the RFDS affected
  list.

Signed-off-by: Pawan Gupta <[e-mail]>
---
Pawan Gupta (10):
      x86/cpu: Prepend 0x to the hex values in cpu_debug_show()
      x86/cpu/topology: Add CPU type to struct cpuinfo_topology
      cpufreq: intel_pstate: Use topology_hw_cpu_type()
      perf/x86/intel: Use topology_hw_cpu_type()
      x86/cpu: Remove get_this_hybrid_cpu_type()
      x86/cpu: Name CPU matching macro more generically (and shorten)
      x86/cpu: Add cpu_type to struct x86_cpu_id
      x86/cpu: Update x86_match_cpu() to also use cpu-type
      x86/bugs: Declutter vulnerable CPU list
      x86/rfds: Exclude P-only parts from the RFDS affected list

 .../admin-guide/hw-vuln/reg-file-data-sampling.rst |   8 --
 arch/x86/events/intel/core.c                       |   3 +-
 arch/x86/include/asm/cpu.h                         |   4 +-
 arch/x86/include/asm/cpu_device_id.h               | 113 ++++++----------
 arch/x86/include/asm/processor.h                   |  11 ++
 arch/x86/include/asm/topology.h                    |   8 ++
 arch/x86/kernel/cpu/common.c                       | 148 +++++++++++----------
 arch/x86/kernel/cpu/debugfs.c                      |   5 +-
 arch/x86/kernel/cpu/intel.c                        |  15 +--
 arch/x86/kernel/cpu/match.c                        |  22 +++
 arch/x86/kernel/cpu/topology_common.c              |  11 ++
 drivers/cpufreq/intel_pstate.c                     |  14 +-
 include/linux/mod_devicetable.h                    |   2 +
 13 files changed, 186 insertions(+), 178 deletions(-)
---
base-commit: 9852d85ec9d492ebef56dc5f229416c925758edc
change-id: 20240617-add-cpu-type-4d5e47efc117

Best regards,

Comments

Dave Hansen Sept. 30, 2024, 3:12 p.m. UTC | #1

On 9/30/24 07:47, Pawan Gupta wrote:
> This series adds support for CPU-type (CPUID.1A.EAX[31-24] on Intel) to
> differentiate between hybrid variants P+E, P-only, E-only that share the
> same Family/Model/Stepping. One of the use case for CPU-type is the
> affected CPU table for CPU vulnerabilities, which can now use the CPU-type
> to filter the unaffected variants.

The series looks fine to me now.  The diffstat is _almost_ even, despite
the series adding functionality.

Acked-by: Dave Hansen <[e-mail]>

Pawan Gupta Sept. 30, 2024, 10:28 p.m. UTC | #2

On Mon, Sep 30, 2024 at 08:12:57AM -0700, Dave Hansen wrote:
> On 9/30/24 07:47, Pawan Gupta wrote:
> > This series adds support for CPU-type (CPUID.1A.EAX[31-24] on Intel) to
> > differentiate between hybrid variants P+E, P-only, E-only that share the
> > same Family/Model/Stepping. One of the use case for CPU-type is the
> > affected CPU table for CPU vulnerabilities, which can now use the CPU-type
> > to filter the unaffected variants.
> 
> The series looks fine to me now.  The diffstat is _almost_ even, despite
> the series adding functionality.
> 
> Acked-by: Dave Hansen <[e-mail]>

Thanks.